Dehydrogenation of propane over a hydrothermal-synthesized Ga2O3-Al2O3 catalyst in the presence of carbon dioxide

Article abstract of DOI:10.1039/c5cy02161h

The hydrothermal synthesis method was adopted to prepare a highly active Ga₂O₃-Al₂O₃ catalyst (GA-HS), which displayed superior catalytic performance for dehydrogenation of propane to propylene in the presence of CO₂ (DHP-CO₂). The highest propane conversion on GA-HS was 35.2%, which was much higher than that from the catalysts prepared using the grind-mixture method (8.7%) or coprecipitation method (26.2%). Moreover, propylene selectivity over the GA-HS catalyst was higher than that over the other catalysts in a period of 9 h of reaction. These catalysts were characterized by N₂ physical adsorption, ICP-AES, XRD, TGA, TEM, SEM, DRIFT, Py-FTIR, NH₃-TPD, XPS, ²⁷Al MAS NMR and ⁷¹Ga MAS NMR techniques. The characterization data indicated that the hydrothermal treatment increased the surface area, expanded the pore size and promoted the formation of more tetrahedral Ga ions and the generation of more medium-strong Lewis acid sites. Furthermore, this catalyst mainly displayed an amorphous sponge-like morphology as well as a new morphology whereby some pieces were covered with amorphous nanoparticles. The superior activity of GA-HS was attributed to the higher surface area of this catalyst and the larger amount of tetrahedral Ga ions (Ga³⁺ and probably Ga^δ+δ < 2) related to the medium-strong Lewis acid sites.

Full text of DOI:10.1039/c5cy02161h

View Article Online
View Journal
Catalysis
Sc ience &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: H. Xiao, J. Zhang,
P. Wang, X. Wang, F. Pang, Z. Zhang and Y. Tan, Catal. Sci. Technol., 2016, DOI: 10.1039/C5CY02161H.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/catalysis

Page 1 of 12
Catalysis Science & Technology
Dynamic Article Links ►
Catalysis Science &
Technology
View Article Online
DOI: 10.1039/C5CY02161H
Cite this: DOI: 10.1039/c0xx00000x
www.rsc.org/xxxxxx
ARTICLE TYPE
Dehydrogenation of propane over hydrothermal synthesized Ga O –
2
3
Al O catalyst in the presence of carbon dioxide
2
3
a,b
a
a,b
a
a,b
a,b
He Xiao, Junfeng Zhang, Peng Wang, Xiaoxing Wang, Fei Pang, Zhenzhou Zhang
a,c
and Yisheng Tan *
5
Received (in XXX, XXX) Xth XXXXXXXXX 20XX, Accepted Xth XXXXXXXXX 20XX
DOI: 10.1039/b000000x
Hydrothermal synthesis method was adopted to prepare a highly active Ga O ꢀAl O catalyst (GAꢀHS),
2
3
2
3
which displayed superior catalytic performance for dehydrogenation of propane to propylene in the
presence of CO (DHPꢀCO ). The highest propane conversion on GAꢀHS was 35.2%, which was much
2
2
1
0
higher than the catalysts prepared using grindꢀmixture method (8.7%) or coprecipitation method (26.2%).
Moreover, propylene selectivity over GAꢀHS catalyst was higher than that over other catalysts in a period
of 9 h reaction. These catalysts were characterized by N physical adsorption, ICPꢀAES, XRD, TGA,
2
2
7
71
TEM, SEM, DRIFT, PyꢀFTIR, NH ꢀTPD, XPS, Al MAS NMR and Ga MAS NMR techniques. The
3
characterization data indicated that hydrothermal treatment increased the surface area, expanded the pore
size and promoted the formation of more tetrahedral Ga ions and generation of more mediumꢀstrong
Lewis acid sites. Furthermore, this catalyst mainly displayed amorphous spongeꢀlike morphology as well
as a new morphology that some pieces were covered with amorphous nanoparticles. The superior activity
of GAꢀHS was attributed to higher surface area of this catalyst and larger amount of tetrahedral Ga ions
1
5
3
+
δ+
(Ga and probably Ga δ< 2) related to mediumꢀstrong Lewis acid sites.
2
0
1. Introduction
activity in the dehydrogenation of ethane to ethylene in the
presence of CO , compared with Cr O and V O catalysts.
Zheng et al. investigated dehydrogenation of propane to
2
2
3
2
5
At present, urgent demand of propylene has drawn academic and
industrial sectors’ attention to deeply investigating the catalytic
dehydrogenation of propane to propylene, for it represents a way
in which more expensive alkenes can be produced economically
36
5
5
6
6
7
7
0
5
0
5
0
5
propylene in the presence of CO2 (DHPꢀCO ) over four
2
polymorphs of gallium oxides. They thought that the superior
performance of βꢀGa O catalyst was ascribed to an abundance of
1
,
2
2
3
2
3
3
4
4
5
0
5
0
5
from lowꢀcost saturated hydrocarbons.
However, the
surface mediumꢀstrong Lewis acid sites related to the unsaturated
dehydrogenation of propane has inherent disadvantages such as
thermodynamic limitations for propane conversion, high energy
requirements ascribed to endothermic reaction and limited
3+
Ga cations and the conjugated effect of proton and oxide. The
optimal propane conversion and propylene selectivity could
achieve to 23% and 95% respectively, at 500 °C with feed gas
3
, 4
catalytic stability in virtue of coke formation, etc.
Recently,
ꢀ1 ꢀ1
mixture (CO /C H molar ratio = 2:1) and flow rate of 3 L g h .
2
3
8
oxidative dehydrogenation of propane with oxygen as an
alternative process for propylene production offers a new
promising technology in term of its energyꢀsaving. By using
oxygen as oxidant, the dehydrogenation can be proceeded well,
but dramatic decrease of propylene selectivity due to excessive
propane oxidation is still a problem. Considering that, CO as a
Based on that, supported gallium oxide catalysts (Ga O /TiO ,
2
3
2
Ga O /Al O , Ga O /SiO , Ga O /ZrO and Ga O /MgO) were
2
3
2
3
2
3
2
2
3
2
2
3
also used for dehydrogenation of propane to propylene with CO
2
37
by Bingjun Xu et al. It was found that Ga O /TiO , Ga O /Al O ,
2
3
2
2
3
2
3
and Ga O /ZrO
showed
a
stronger ability for the
2
2
3
2
2
dehydrogenation of propane than Ga O /SiO and Ga O /MgO,
2
3
2
2
3
mild oxidant is used to develop a safer and more economical
process for propane dehydrogenation. With introduction of
due to the abundant mediumꢀstrong Lewis acid sites on these
catalysts’ surface and special interactions between the support
and Ga O . Ga O /TiO showed the highest activity, and the
5
ꢀ7
CO , the dehydrogenation efficiency is increased because of the
2
2
3
2
3
2
new additional production process of propylene: a direct surface
redox mechanism in which the catalyst undergoes reduction (by
propane) and reoxidation (by carbon dioxide) cycles
propane conversion over this catalyst highly achieved to 39% at
00 °C with feed gas mixture (CO /C H molar ratio= 2:1) and
6
8
ꢀ10
2
3
8
or a more
ꢀ1 ꢀ1
30
flow rate of 6 L g h . Michorczyk et al. prepared ordered
mesoporous Ga O and Ga O ꢀAl O by nanocasting, which were
complex reaction pathway involving a simple dehydrogenation
11ꢀ13
2
3
2
3
2
3
followed by the reverse water gas shift (RWGS) reaction.
Ga O ꢀcontaing catalysts have been considered for alkane
used as effective catalysts for DHPꢀCO . The results revealed that
2
2
3
the pure mesoporous gallium and galliumꢀaluminum oxides
oxidative dehydrogenation processes due to their relatively good
catalytic performance.
According to their result, the commercial Ga O showed higher
(Ga/Al = 4:1) exhibited a promising catalytic performance in the
1
4ꢀ34
35
This was verified by Nakagawa et al.
DHPꢀCO . Over the most active materials, propylene was
2
2
3
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 | 1

Catalysis Science & Technology
Page 2 of 12
View Article Online
DOI: 10.1039/C5CY02161H
produced with the yield of 10ꢀ18% and high selectivity of 91ꢀ95% h under static air in a muffle oven, the catalyst was obtained and
during 4 h on stream at 550 °C with feed gas mixture (CO /C H
8
defined as GAꢀGM.
2
3
ꢀ
1
ꢀ1
molar ratio= 5:1) and flow rate of 9 L g h . Moreover, Chen et
Coprecipitation method: the Ga O –Al O catalyst was
2
3
2
3
11, 12
40
al.
studied the effect of preparation method and treatment 60 prepared through an alcoholic coprecipitation pathway. In a
5
0
5
0
5
0
5
0
process on the activity and stability of Ga O ꢀcontaing catalysts
typical synthesis, 0.01 mol Ga(NO3)3 and 0.02 mol Al(NO3)3
(Ga/Al = 1:2) were completely dissolved in 40 mL ethanol. Then,
40 mL concentrated aqueous ammonia (28 wt%) was added
dropwise to above ethanol solution. The resulting suspension was
2
3
for the dehydrogenation of propane. The results suggested that
3
+
larger amount of surface Ga sites with weak Lewis acidity were
presented on galliaꢀalumina solid solution prepared using
coprecipitation method than that on the Ga O –Al O oxide 65 filtrated and washed by ethanol. After that, the cake was dried at
2
3
2
3
1
1
2
2
3
3
4
catalysts prepared using physical mixture method, thereby
leading to its better performance in propane dehydrogenation
120 °C for 8 h in constant temperature oven, followed by
calcination at 550 °C for 5 h (from room temperature to 550 °C at
a ramp rate of 3 °C/min). The obtained catalyst was defined as
GAꢀCP.
with CO . It was found that the initial conversion of propane and
2
selectivity of propylene over Ga Al O were 49.7% and 91.7%
8
2
15
respectively, at 500 °C with feed gas mixture (CO /C H molar
70
Hydrothermal synthesis method: 0.01 mol Ga(NO ) and 0.02
2
3
8
3 3
ꢀ
1
ꢀ1
ratio = 2:1) and flow rate of 3 L g h . For the sake of deeply
mol Al(NO ) (Ga/Al = 1:2) were completely dissolved in 40 mL
3
3
3
6ꢀ38
understanding this reaction, many researchers
drew their
ethanol. Then, 40 mL concentrated aqueous ammonia (28 wt%)
was added dropwise to above ethanol solution. The resulting
suspension was transferred into 200 ml Teflon cup inserted in
3
+
attention to investigating the influence of surface Ga in Ga O ꢀ
2
3
containg catalysts on dehydrogenation of propane. And their
results suggested that hydrocarbon activation proceeded on lowꢀ 75 stainless steel autoclave, then heated to 170 °C from room
3
+
coordinated Ga
species,
cations, accordingly forming propylꢀGa
and the dehydrogenation products were then
generated via subsequent decomposition of the resulting propylꢀ
temperature at a ramp rate of 10 °C/min and kept for 24 h. After
that, the crystalline product was washed repetitively with ethanol,
followed by drying at 120 °C for 8 h in constant temperature
oven, and calcined at 550 °C for 5 h under static air like that of
3
6, 37
3+
Ga species. Specifically speaking, the lowꢀcoordinated Ga sites
over gallium oxide were highly effective in the heterolytic 80 coprecipitation method. The obtained catalyst was defined as
dissociative adsorption of propane, because of the very strong
polarizability of the C–H bonds resulted from perturbation of
GAꢀHS.
Compared with GAꢀHS, pure Ga (GꢀHS) and Al (AꢀHS) were
3
+
39
propane on the lowꢀcoordinated Ga sites. It’s obvious that
more active catalysts for propane dehydrogenation can be
prepared using hydrothermal synthesis method. Moreover,
Ga O ꢀAl O materials with different Ga/Al molar ratios (1:1 and
2
3
2
3
obtained by getting a higher population of lowꢀcoordinated 85 2:1) were also prepared by grindꢀmixture method, coprecipitation
3+
surface Ga sites in the Ga O ꢀcontaing catalysts.
method and hydrothermal synthesis method, defined as G1A1ꢀ
GM, G2A1ꢀGM, G1A1ꢀCP, G2A1ꢀCP, G1A1ꢀHS and G2A1ꢀHS
respectively.
2
3
In this work, a facile hydrothermal synthesis method was
adopted to prepare a Ga O ꢀAl O catalyst for DHPꢀCO . The asꢀ
2
3
2
3
2
prepared catalyst displayed better catalytic performance than the
catalysts prepared using grindꢀmixture method or coprecipitation
2
.2 Catalyst characterization
method. We have inꢀdetail rationalized the results by various 90 The textural properties of catalyst were measured by N
characterizations such as N physical adsorption, ICPꢀAES, XRD,
sorption at 77K using a Tristar 3000 machine. Surface areas were
physical
2
2
2
7
TEM, TGA, SEM, DRIFT, PyꢀFTIR, NH ꢀTPD, XPS, Al MAS
NMR and Ga MAS NMR techniques, and aimed to unveil the
calculated by the BET method and microꢀ, mesoꢀ, and macropore
volumes were calculated by the tꢀplot method.
3
71
correlations between the structure and physical properties of
The powder Xꢀray diffraction (XRD) patterns were measured
Ga O ꢀAl O
3
dehydrogenation
catalyst and its activity for propane 95 on a Rigaku MiniFlex II Xꢀray diffractometer using CuKα
radiation. The anode was operated at 40 kV and 40 mA. The 2θ
angles were scanned from 10° to 70°.
2
3
2
Diffuse reflectance infrared spectra (DRIFT) measurements
2
. Experimental section
were performed using a Bruker Tensor 27 instrument with a MCT
2
.1 Catalyst preparation
−1
1
00 detector (64 scans, 4 cm ). Simply, 20 mg of catalyst was put it
Herein, Ga O ꢀAl O catalysts as well as the simple oxide of
Ga O and Al O were prepared using different methods as
2 3 2 3
in an infrared cell with KBr windows for inꢀsitu treatments. The
DRIFT spectra were recorded after treating the catalyst at 300 °C
for 1 h with a flow of argon.
2
3
2
3
45
50
55
follows.
Ga O and Al O can be obtained from Ga(NO3)3 and
Pyridineꢀadsorbed Fourier Transform Infrared Spectroscopy
Al(NO ) respectively through the same treatment of drying at 105 (PyꢀFTIR) was used to determine the amount of Brønsted acid
2
3
2
3
3
3
1
20 °C for 8 h in constant temperature oven and subsequent
sites and Lewis acid sites using Bruker Tensor 27 equipment. 20
mg of catalyst was pressed into a regular wafer (R = 1.3 cm) and
then put in an infrared cell. The infrared spectrum was recorded
after sample treatment at 400 °C for 2 h under vacuum, wherein
calcinations at 550 °C for 5 h (from room temperature to 550 °C
at a ramp rate of 3 °C/min) under static air in a muffle oven.
Grindꢀmixture method: 0.01 mol Ga(NO3)3 and 0.02 mol
Al(NO ) (Ga/Al = 1:2) as the precursors were firstly mixed and 110 this spectrum was used as background for the adsorbed pyridine
3
3
ꢀ
2
ground, followed by drying at 120 °C for 8 h in constant
temperature oven. The resulting mixture was heated to 550 °C
from room temperature at a ramp rate of 3 °C/min, and kept for 5
experiments. Pyridine was then adsorbed to a 5.0×10 Pa
equilibrium pressure at 40 °C. FTIR spectra were recorded after
consecutive evacuation at 400 °C.
The temperatureꢀprogrammed desorption of ammonia (NH3ꢀ
2
| Catal. Sci. Technol. [2015], [vol], 00–00
This journal is © The Royal Society of Chemistry [2015]

Page 3 of 12
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY02161H
ꢆꢇꢈꢉꢊꢈꢋꢌꢍ,ꢎꢌꢏꢇꢈꢉꢊꢈꢋꢌꢍ,ꢊꢐꢑ
ꢒ
TPD) was used to test the amount and strength of the Lewis acid
or Brønsted acid sites of asꢀprepared catalysts in TPꢀ5080
chemisorption instrument. The catalyst (100 mg) was preꢀtreated
Xꢀꢁꢂꢀꢃꢄꢅ
=
× 100%
(1)
(2)
ꢇ
ꢈꢉꢊꢈꢋꢌꢍ,ꢎꢌ
ꢇ
ꢈꢉꢊꢈꢖꢗꢍꢌꢍ
Sꢀꢁꢂꢀꢓꢔꢅꢄꢅ =ꢕ
× 100%
at 500 °C under a flow of N (30 ml/min) for 2 h and then cooled
2
ꢇꢈꢉꢊꢈꢋꢌꢍ,ꢎꢌꢏꢇꢈꢉꢊꢈꢋꢌꢍ,ꢊꢐꢑ
5
0
5
0
5
0
5
0
down to 100 °C. Then NH was introduced into the flow system.
The TPD spectra were recorded after about 1h evacuation time
until the baseline of TCD signal is steady at a ramp rate of
3
6
0
of moles of propane in the inlet and outlet gas phases,
respectively, and Npropylene,out is the number of moles of product
propylene.
1
0 °C/min from 100 °C to 600 °C.
Xꢀray photoelectron spectroscopy (XPS) was used to analyse
Regeneration conditions: T = 600 °C; air or pure CO (99.99%);
1
1
2
2
3
3
4
the change of surface composition measured by AXIS ULTRA
DLD equipment. The binding energy values were corrected for
charging effect by referring to the adventitious C1s line at 284.5
eV.
The chemical compositions (Al and Ga) of asꢀprepared samples
were determined by an inductively coupled plasmaꢀatomic
emission spectrometer (ICPꢀAES) method using Thermo iCAP
2
flow rate = 30 ml/min; treatment time = 10 h.
65 3. Results and discussion
3
.1 Catalyst test
The conversion of propane and the selectivity of propylene over
these asꢀprepared samples as a function of time on stream are
6
300 equipment.
71
shown in Fig. 1. Compared with single Al O , the activity of
Ga MAS NMR measurements were performed on Bruker
2
3
7
1
7
7
8
8
9
9
0
5
0
5
0
5
Ga O ꢀcontaing catalysts are remarkably superior for propane
2 3
AVANCE III 600 MHz equipment. Ga MAS NMR spectra
were obtained by a single pulse length of Π/6, and Ga chemical
shifts were referenced to Ga(NO ) .
dehydrogenation (Fig. 1A), indicating that the presence of Ga
species is necessary to generate active sites for propane
3
3
35
27
dehydrogenation. Especially on GAꢀHS, the initial propane
conversion is the highest (35.2%), which is much higher than that
Al MAS NMR measurements were performed on Bruker
2
7
AVANCE III 600 MHz equipment. Al MAS NMR spectra were
obtained by a single pulse length of Π/6, and Al chemical shifts
were referenced to Al(NO ) .
on Ga O (5.7%), GAꢀGM (8.7%) and GAꢀCP (26.2%). From
2
3
Fig. 1B, it is noticeable that propylene selectivity over these
Ga O ꢀcontaining catalysts achieve to about 95.0%, which are
3
3
Transmission electron microscopies (TEM) were performed
using a JEMꢀ2100 transmission electron microscope.
Scanning electron microscopy (SEM) images were obtained
for morphologic identification using a JSMꢀ7001F.
Pulsed reaction technique was used to measure the activities
of these catalysts and carried out on chemisorption instrument
2
3
obviously higher than that on single Al O3 (only 73.2%).
2
Propylene selectivity over Ga O and GAꢀGM decrease by 13.7%
2
3
and 26.5 % respectively after 9 h reaction, while that over GAꢀ
HS and GAꢀCP almost remain unchanged during the running
period accompanied by the decline of propane conversion.
Seemingly, the propylene selectivity over GAꢀHS is slightly
higher than that over GAꢀCP. These test results suggest that the
(
TPꢀ5080) and OMNI star. Transient responses of m/z = 44 (C H8
3
and CO ) and m/z = 44, 43 (C H ) over asꢀprepared catalysts was
2
3
8
activities of Ga O ꢀcontaining catalysts are strongly affected by
tested by an onꢀline mass spectrometer against a pulsed
introduction of C H under steady flow of mixture gas (10% CO
2
2
3
the preparation method and treatment process. And our adopted
hydrothermal synthesis method is remarkably favourable for
preparing the catalyst for efficient dehydrogenation of propane, in
comparison with the grindꢀmixture method or coprecipitation
3
8
and 90% Ar). Reaction conditions: catalyst = 100 mg; CO and Ar
2
ꢀ
1
carrier = 30 mlmin ; C H = 1 ml; furnace temperature = 550°C.
3
8
Thermogravimetric analysis (TGA) was conducted on a
Rigaku TG analyzer to determine the amount of coke deposited
on the catalyst after the reaction. 10 mg of sample was heated
from room temperature to 800 °C at a heating rate of 10 °C/min
in flowing air.
method. In addition, pure Ga O (GꢀHS) and Al O (AꢀHS) are
2
3
2
3
prepared using hydrothermal synthesis method in comparison
with GAꢀHS. The initial propane conversions on GꢀHS and AꢀHS
are 21.4% and 0.8% (Fig. S1a) respectively, much lower than that
on GAꢀHS, implying that single Ga O or Al O component is
2
3
2
3
2
.3 Catalyst evaluation
The catalyst test was carried out in a horizontal quartz tube fixedꢀ
bed reactor. Reaction conditions: T = 550 °C; P = 100 kPa; mcat.
50 mg; the feed gas mixture C H /CO /N with the molar ratio
indifferent for propane dehydrogenation but their synergistic
effect is active due to forming special structure. Besides, the rapid
decreases of propylene selectivity on GꢀHS and AꢀHS (Fig. S1b)
are seen with increase of time on stream in comparison with GAꢀ
4
5
5
5
0
5
=
1
3
3
8
2
2
−1
of 1:3:27; total flow rate = 15 cm ·min .
The products were analysed by offꢀline gas chromatograph 100 ratios (1:1 and 2:1) are also investigated. From evaluation results
HUAAI GC 9560) equipped with a FID with Al O3 packed
of these catalysts, it’s found that the conversions of propane on
the Ga ꢀAl catalysts are enhanced with the increase of
Ga/Al molar ratio (Fig. S1a). Wherein, the Ga ꢀAl prepared
using hydrothermal synthesis method still shows the superior
to
those catalysts prepared using grindꢀmixture method and
HS. Moreover, Ga O ꢀAl O materials with different Ga/Al molar
2 3 2 3
(
2
column, and offline gas chromatograph (East & West GC 4000A)
equipped with a thermal conductivity detector (TCD) with carbon
molecular sieves packed column, respectively. The propane
2
O
3
O
2 3
O
3
O
2 3
2
conversion and propylene selectivity were calculated using the 105 activity for propane dehydrogenation in the presence of CO
2
following eqs (1) and (2).
Where Xpropane is the conversion of propane, Spropylene is the
selectivity of propylene, Npropane,in and Npropane,out are the numbers
coprecipitation method with the same Ga/Al molar ratio. The
propylene selectivity over the hydrothermally prepared catalyst
achieves to about 95.0 % during the whole test period like the
10 coprecipitatedꢀprepared catalyst (Fig. S1b), which is higher than
1
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 | 3

Catalysis Science & Technology
Page 4 of 12
View Article Online
DOI: 10.1039/C5CY02161H
that over the catalyst prepared by grindꢀmixture method.
In order to study the effect of CO on this reaction, the catalytic
2
tests without CO2 are performed over these Ga O ꢀAl O
3
2
3
2
1
2
2
3
3
4
4
5
5
6
6
5
0
5
0
5
0
5
0
5
0
5
catalysts for comparison (Fig. S2 and Table S1). The initial
conversions of propane on these Ga O ꢀAl O catalysts without
2
3
2
3
CO are higher than that with CO , indicating that the presence of
2
2
carbon dioxide is not positive to promote the dehydrogenation of
37
propane as expected. Xu et al. considered that the negative
effect was been attributed to the greatly reduced propane
adsorption capacity due to the competition of CO . However,
2
compared with catalytic tests in presence of CO , rapid
2
deactivation of these Ga O ꢀAl O catalysts in the absence of CO
2
3
2
3
2
is observed. And propane conversions on these Ga O ꢀAl O
2
3
2
3
catalysts testing in presence of CO are higher than that in the
2
absence of CO after 9 h on stream. It can be indicated that the
2
presence of CO2 is beneficial for improving durability of the
catalysts. The enhanced stability of these Ga O ꢀAl O catalysts
2
3
2
3
in the presence of CO may be due to their low coking
2
12
tendency. In addition, the primary product (propylene) and byꢀ
products (methane, ethylene, ethane, etc.) are formed during the
dehydrogenation process of propane regardless of the presence of
CO . However, it is worthy to note that the selectivity to byꢀ
2
products over these Ga O ꢀAl O catalysts in the presence of CO
2
2
3
2
3
is lower than that in the absence of CO . It is thought that CO as
2
2
a mild oxidant can promote the dehydrogenation of propane
preferentially and weaken the cracking and hydrocracking
reactions somewhat, thereby increasing the selectivity of target
Fig. 1 Propane conversion (A) and selectivity of propylene (B) as a
function of time on stream for asꢀprepared samples. Reaction conditions:
T = 550 °C; P = 100 kPa; mcat. = 150 mg; the feed gas mixture
product propylene compared with that in absence of CO . With
2
5
the observation of higher conversion of CO and lower molar
2
C
3
H
8
/CO
2
/N
2
with the molar ratio of 1:3:27; total flow rate = 15
ratio H /CO over GAꢀHS (Table S1), it can be further confirmed
2
that GAꢀHS has superior activity to GAꢀGM and GAꢀCP in the
present reaction. The catalytic tests of GAꢀHS after regeneration
3
−1
cm ·min .
with air or CO are performed (Fig. S3). The comparative results
2
of regeneration with O and CO indicate that both agents are
2
2
effective in coke removing. But propane conversion and
propylene selectivity on regenerated GAꢀHS with CO after 9h
2
reaction on stream are much lower than that with air, implying
that the CO can partly remove the coke through Boudouard
2
37
reaction.
Considering the distinction of test conditions (such as reaction
temperature, GHSV and CO /C H molar ratio of feed gas
2
3
8
mixture, etc.), it is difficult to make a certain conclusion that
which preparation method is more preferential by comparing the
activities of the Ga O ꢀcontaining catalysts prepared by different
2
3
researchers. But it is widely accepted that the presence of Lewis
3+
acid sites related to Ga ions in tetrahedral sites, which is
obtained through special treatment process such as hydrothermal
synthesis, is favourable for propane dehydrogenation. However,
propane conversion and propylene selectivity over GAꢀHS are
higher than that over these catalysts prepared using grindꢀmixture
method or coprecipitation method at the same test conditions,
indicating that the hydrothermal synthesis can improve the
activity of Ga O ꢀAl O catalysts greatly. In this work, we focus
2
3
2
3
on the effect of preparation method on the activity of Ga O ꢀ
2
3
Al O catalysts for DHPꢀCO .
2
3
2
Propane conversions over these Ga O ꢀcontaining samples all
2
3
10
decrease as a function of reaction time. Wherein, it is thought that
this gradual deactivation of the catalysts is caused by the coke
Fig. 2 TG (A) and DTA (B) profiles of spent Gaꢀcontaining catalysts.
| Catal. Sci. Technol. [2015], [vol], 00–00
4
This journal is © The Royal Society of Chemistry [2015]

Page 5 of 12
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY02161H
2
5, 36
deposition on the surface of these catalysts.
Based on that,
on stream. In addition, DTA profiles of the spent Gaꢀcontaining
(Fig. 2B) show that both spent GAꢀGM and spent Ga O only
display one peak at 460 °C and 426 °C respectively,
coke deposition on the spent Gaꢀcontaining samples is quantified
by TGꢀDTA as show in Fig. 2A. The weight loss of these spent
2
3
samples decreases in the order: GAꢀHS > GAꢀCP > GAꢀGM > 40 corresponding to one type of carbon species. But the spent GAꢀ
5
0
5
0
5
Ga O . Spent GAꢀHS displays maximum weight loss compared
CP shows two peaks at 454 °C and 479 °C assigned to two
different types of carbon species. Spent GAꢀHS displays a similar
DTA profile to the spent GAꢀCP, but with higher amounts of
carbon species. Nevertheless, it deserves further study to
2
3
with the Ga O ꢀAl O catalysts prepared using grindꢀmixture
2
3
2
3
method or coprecipitation method, demonstrating its high coke
1
9, 30
performance. Published work
reported that the high catalytic
performance of Gaꢀcontaining materials is attributed to the 45 understand the details about the mechanism of formation of
1
1
2
2
surface gallium sites with Lewis acidity, which are also the coke
deposition sites. For the present catalysts, GAꢀHS should possess
more Lewis acid sites than the others based on its relatively good
performance (as shown in Fig. 1A). That implies that there are
more coke deposition sites existed on GAꢀHS. Moreover, the
different carbon species.
.2 Catalyst composition and textural properties
The pore volume, pore size and BET surface areas (SBET) of
different catalysts are obtained based on sorption isotherms of N
ꢀsorption isotherms of these catalysts
(Fig. S4) are a type ІV with a steep increase around P/P = 0.4,
indicating that these samples belong to typical mesoporous
3
2
initial conversion of propane (Fig. S2) and the initial selectivity 50 condensation at 77 K. N
2
of propylene (Table S1) over these Ga O ꢀAl O catalysts in the
0
2
3
2
3
absence of CO are much higher. This result suggests that the
2
41, 42
concentration of propylene in the absence of CO in the initial
materials.
From the profiles of pore size distributions in Fig.
2
step is also high. Because the propylene is more reactive than
propane, its higher concentration in reaction zone leads probably 55 pore size distribution with average pore diameter 7.6 nm and 7.9
3, it can be seen that GAꢀGM and GAꢀCP have a very narrow
to higher coke amount, which also contributes to fast deactivation
of catalyst in the dehydrogenation of propane in terms of
nm respectively, while GAꢀHS has a broad pore size distribution
with average pore diameter 12.5 nm. This suggests that special
pore structure can be produced after the hydrothermal treatment.
The BET surface area and total volumes of the catalysts are
1
7, 30
substratesꢀproducts reactivity.
Thus, the coke is easily formed
on GAꢀHS catalyst during DHPꢀCO2 process due to its high
activity. In addition, the coke also has significant effects on the 60 calculated and the results are shown in Table 1. The BET surface
durability of these catalysts. As our observation, the propane
conversion decrease dramatically with increase of time
2
3
area and total volumes of GAꢀHS are 234 m /g and 0.730 cm /g
respectively, higher than that of GAꢀGM, GAꢀCP, pure Ga
and pure Al O . In combination with the above test results,
O
3
2
2
3
3
0
Table 1 Textural properties of different samples
65 Table 2 Textural properties of different spent and regenerated Ga
2
O ꢀ
3
Al
2
O
3
catalysts
a
a
b
total
3
b
S
BET
2
S
micro
2
S
ext
V
V
micro
3
V
ext
(cm /g)
2
3
Sample (m /g) (m /g)
( m /g ) (cm /g) (cm /g)
Al
Ga
2
O
3
68.8
29.2
15.7
5.9
13.7
32.0
22.0
53.1
23.3
26.2
188
212
0.075
0.129
0.076
0.438
0.730
0.007
0.003
0.001
0.013
0.008
0.068
0.126
0.075
0.425
0.722
a
a
b
b
S
BET
2
S
micro
2
S
ext
2
V
total
3
V
micro
3
V
ext
3
D
pore
2
O
3
Sample (m /g) (m /g) ( m /g ) (cm /g) (cm /g) (cm /g) (nm)
GAꢀGM 39.9
GAꢀCP
GAꢀHS
GAꢀGM 73.0
74.6)
GAꢀCP 123.8
(161.4) (8.4) (153.0) (0.235) (0.002)
GAꢀHS 160.0 3.4 156.6 0.187 0.004
5.7
(5.5) (69.1) (0.107) (0.002)
12.2 111.6 0.141 0.008
67.3
0.089
0.003
0.086
(0.105) (5.7)
0.133 4.5
(0.233) (5.8)
0.183 4.6
4.9
220
234
(
a
b
0
determined by tꢀmethod; determined by Volume adsorbed at p/p =0.97
(190.7) (4.3) (186.4) (0.322) (0.004)
(0.318) (6.7)
a
b
0
determined by tꢀmethod; determined by Volume adsorbed at p/p =0.97
The value outside and inside the bracket are the data of the BET surface
areas, pore volumes and pore diameter obtained after reaction and after
oneꢀround regeneration with air (30 ml/ min) respectively.
70
Table 3 Compositions and numbers of acid sites in asꢀprepared catalysts
Tc
T
d
e
Ga/Al molar ratio
Ga
Ga total
NH
3
desorbed
Sample
Al
Ga
a
b
ꢀ1
Bulk Surface
(%)
(%)
(mmol g cat
0.079
0.047
)
2
O
3
/
/
/
/
2
O
3
/
/
9.5
9.5
GAꢀGM
GAꢀCP
GAꢀHS
0.587
0.536
0.576
0.505
0.920
1.11
37.7
78.3
85.5
13.9
27.3
31.2
0.073
0.141
0.168
a
b
c
75
determined by ICPꢀAES method; determined by XPS; determined by
7
1
d
Ga MAS NMR; Ga total (IV) % = nGa/(nGa + nAl) x Ga (IV)%, nGa/(nGa + nAl
)
e
is the bulk molar ratio; determined by NH
3
ꢀTPD.
3
5
Fig. 3 Pore size distributions for different catalysts.
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 |
5

Catalysis Science & Technology
Page 6 of 12
View Article Online
DOI: 10.1039/C5CY02161H
45
Fig. 4. XRD profiles of asꢀprepared samples (a) Ga
GAꢀCP; (d) GAꢀHS; (e) Al
2 3
O ; (b) GAꢀGM; (c)
71
Fig. 5 Ga MAS NMR profiles of different catalysts. (a) GAꢀGM; (b)
2
3
O .
GAꢀCP; (c) GAꢀHS.
5
0
5
0
5
0
5
0
it is thought that active Ga species can be more easily dipersed on
GAꢀHS due to its high BET surface and volumes than that of GAꢀ
GM and GAꢀCP, thereby leading to its good performance
probably. Moreover, the higher micropore surface area and
micropore volumes of GAꢀCP than that of GAꢀHS and GAꢀGM
means that using a coprecipitation method can lead to the
preferential formation of micropores over GAꢀCP than other
catalysts. Based on the larger number of mesꢀ and macroꢀpores in
GAꢀHS than GAꢀCP and GAꢀGM, the conclusion can be made
that hydrothermal synthesis promotes the formation of mesꢀ and
macroꢀpores over the Ga O ꢀAl O catalyst, but the
1
1
2
2
3
3
4
2
3
2
3
coprecipitation method is in favor of formation of micropores.
To clarify the thermal stability of the catalyst prepared by
hydrothermal method, the textural properties of the samples after
DHPꢀCO test and regeneration with air have been characterized
2
by low N adsorption and the data is listed in Table 2. Compared
2
with the fresh samples, the spent GAꢀCP and spent GAꢀHS both
2
show obvious decrease in the SBET (decrease by 96.2 m /g and
2
3
7
4.0 m /g) and pore volume (decrease by 0.297 m /g and 0.543
3
m /g) respectively, whereas the spent GAꢀGM shows some
increase in the S
.014 m /g) due to cumuli of coke deposition. After regeneration
2
and pore volume (incease by 33.1 m /g and
BET
3
0
with air, pores of these spent Ga O ꢀAl O catalysts are recovered
2
3
2
3
partly. And the regenerated GAꢀHS still shows the highest SBET
and pore volume among these regenerated samples, indicating
that the hydrothermal synthesis method is superior to other
preparation methods in this work.
Fig. 6 Ga 3d XPS (A) and Ga 2p1/2 XPS (B) profiles of different catalysts.
(a) GAꢀGM; (b) GAꢀCP; (c) GAꢀHS.
5
0
Fig. 4 illustrates the XRD patterns of these Ga O ꢀAl O
3
2
3
2
catalysts, along with simple Ga O and Al O . The diffraction
2
3
2
3
are formed in physical mixture process. Yet GAꢀCP and GAꢀHS
don’t exhibit typical characteristics of simple γꢀGa O and/or γꢀ
lines of the asꢀprepared samples are very broad, demonstrating a
low crystallinity of all the samples. This low crystallinity is a
common feature of metastable γꢀvariety of both alumina and
2
3
Al O like GAꢀGM. d spacing corresponds to cubic lattice
(440)
2
3
parameter a , which can be used to distinguish the cubic spinelꢀ
12
0
gallia polymorphs. The asꢀprepared Ga O reveals the typical
2
3
55
type structure (galliaꢀalumina solid solution) by the location of
characteristics of γꢀGa O inferred from the 2θ angles (30.8°,
40
2
3
(
440) peak. It can be seen that the (440) peaks of γꢀGa O and γꢀ
2
3
3
6.2° and 64.2° in curve a) of diffraction peak, while simple
Al O are located at 64.2° and 67.1° respectively. The (440) peaks
2
3
Al O suggests the typical characteristics of γꢀAl O based on the
2
3
2
3
of GAꢀCP and GAꢀHS lie between that of γꢀGa O and γꢀAl O ,
2
3
2
3
2
θ angles (45.4°, 60.3° and 67.1° in curve e) of the diffraction
indicating that the cubic spinelꢀtype structures are formed on GAꢀ
peak. GAꢀGM closely resembles γꢀGa O and γꢀAl O in the light
43
2
3
2
3
60
CP and GAꢀHS. In comparison with GAꢀCP, the (440) peak of
of diffraction maxima (curve b), indicating that no new structures
GAꢀHS shifts to high 2θ angle due to preparation method. This
6
| Catal. Sci. Technol. [2015], [vol], 00–00
This journal is © The Royal Society of Chemistry [2015]

Page 7 of 12
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY02161H
suggests that different structure presents on GAꢀCP and GAꢀHS
in despite of the presence of similar cubic spinelꢀtype structures
on these two catalysts, which results in their activity differences 30 hydride and gallium alkoxide species as step (3), and the
dehydrogenation. Propane can heterolytically dissociate on lowꢀ
3
+
coordinated surface Ga sites in gallium oxide, forming gallium
for propane dehydrogenation.
alkoxides then decompose further to the dehydrogenation
products as step (4) and (5) as follows:
5
3.3 Coordination state and oxidation state of Ga
ꢀ
+
H
C H
3
7
3
+
71
3+
2
3+
3+
2−
3+
To reflect the coordination state of Ga ion, the Ga MAS NMR
chemical shift measurements are conducted and the spectra of
different catalysts are shown in Fig. 5. The two asymmetric bands
Ga –O —M + C H → Ga –O –M (M = Ga, Al) (3)
3 8
ꢀ
+
ꢀ
+
35
H
C H
H
H
3
7
3
+
2
3+
3+
2−
3+
Ga –O —M + C H → Ga –O –M + C H6
(4)
3
8
3
T
within ꢀ100 ppm ~ 300 ppm are assigned to tetrahedral Ga (Ga )
and
O
14, 40, 44ꢀ46
ꢀ
+
1
1
2
2
0
5
0
5
ions and octahedral Ga (Ga ) ions, respectively.
Peak
H
H
T
O
3+
2
3+
3+
2−
3+
deconvolution of the NMR Ga and Ga bands are also done to
Ga –O —M → Ga –O –M + H2
(5)
provide more information about the Ga distribution (among 40 When CO2 is introduced into the propane dehydrogenation
tetrahedral and octahedral sites) in the structure of these catalysts.
reaction, chemisorbed H2 generates from the step (5) can be
removed via an alternative route as step (6):
The deconvoluted results in Table 3 suggest that most of Ga ions
O
ꢀ
+
in GAꢀGM and Ga O occupy octahedral Ga sites while that in
H
H
2
3
T
3+
2
3+
3+
2−
3+
GAꢀCP and GAꢀHS are mainly located at tetrahedral Ga sites.
Especially for GAꢀHS, the Ga ions percent achieves to 85.5%. It
Ga –O —M + CO → Ga –O –M + CO + H O (6)
2 2
T
45
3
+
indicates that hydrothermal synthesis method promotes the
It’s apparent that the more lowꢀcoordinated surface Ga sites are
obtained in the Ga O ꢀcontaing catalysts, the better catalytic
T
formation of tetrahedral Ga ions. Besides, the content of total
2
3
T
Ga ions is also calculated in combination with bulk content of
activity in propane dehydrogenation is. Apart from that, the
3+
27
the Ga obtained by ICPꢀAES. The results show that the content of
coordination state of the Al ion is also investigated by Al
T
total Ga ions of GAꢀHS is higher than GAꢀGM and GAꢀCP. 50 MAS NMR and the results are shown in Fig. S5 and the
3
6, 37
3+
Already published research
indicated that the lowꢀ
percentage of Al ions are listed in Table S2. The results suggest
T
3+
coordinated Ga cations contributed to the activity of Ga O ꢀ
that the percentage of Al ion in GAꢀHS is higher than that in
2
3
containg catalysts for DHPꢀCO . And it is consistent with our
results that the content of Ga cations has a positive relationship
GAꢀGM and GAꢀCP, indicating that hydrothermal synthesis
2
T
with activity of Ga O ꢀcontaing catalysts for propane
2
3
5
5
Fig. 7 TEM images of different asꢀprepared catalysts: (a) Al
2
O
3
; (b) Ga
2
O
3
; (c) GAꢀGM; (d) GAꢀCP; (e, f) GAꢀHS.
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 | 7

Catalysis Science & Technology
Page 8 of 12
View Article Online
DOI: 10.1039/C5CY02161H
T
37
δ+
promotes the formation of more tetrahedral Al ions as well as
tetrahedral Ga ions.
reaction. The reduced gallium ions (Ga species δ< 2) are also
60 thought to have high dehydrogenation efficiency. The galliumꢀ
hydrogen bond can be formed on these reduced gallium ions by
heterolytic hydrogen dissociation, and can be stabilized on the
T
XPS is an effective technique to characterize the surface
4
7ꢀ52
5
0
5
0
5
0
5
0
5
0
5
situation of catalysts.
Oxidation states and quantity of the
elements that are present within the top 1ꢀ12 nm of the sample
surface can be obtained by this technique. The surface Ga/Al
21
support surface. Therefore, dehydrogenation proceeds on the
supported gallium oxide catalysts probably through a heterolytic
ratios measured by XPS and the bulk Ga/Al ratios determined by 65 dissociation reaction pathway instead of redox mechanism.
ICPꢀAES are shown in Table 3. The surface Ga/Al ratio of GAꢀ
GM is lower than bulk Ga/Al ratio, while the surface Ga/Al ratios
of GAꢀCP and GAꢀHS are higher than their bulk Ga/Al ratios.
This indicates that the coprecipitation synthesis and hydrothermal
synthesis could make more Ga species expose on the surface of
Wherein, it should be pointed out that the further work needs to
δ+
1
1
2
2
3
3
4
4
5
5
be done to clarify the mechanism of conversion of Ga species
3+
(δ< 2) to Ga species with and without CO2.
3
.4 Morphologies of as-prepared samples
Ga O ꢀAl O catalysts. This is also verified by SEMꢀEDS results 70 In order to reveal the morphologies of these asꢀprepared samples,
2
3
2
3
(Fig. S6), which show that the surfaces of GAꢀCP and GAꢀHS are
remarkably enriched in Ga, compared with GAꢀGM. The Ga 3d
and Ga 2p1/2 XPS spectra of these catalysts are displayed in Fig. 6.
From the deconvoluted spectra on the Ga 3d region (Fig. 6A), the
TEM is applied and the results are shown in Fig. 7. Al
2 3
O shows
smooth pieces (Fig. 7a) while Ga takes on an irregular and
O
3
2
segregated morphology (Fig. 7b). GAꢀGM as shown in Fig. 7c
mainly contains two kinds of morphologies. Wherein, it is
observation is that only two peaks appear in the spectra of GAꢀ 75 thought that they are ascribed to Ga O and Al O respectively, in
2 3 2 3
GM and GAꢀCP at about 24.1 and 20.6 eV respectively, which
combination with Fig. 7a, 7b. While GAꢀCP (Fig. 7d) just
exhibits a unified distinct spongeꢀlike morphology probably
ascribed to galliaꢀalumina solid solution (cubic spinelꢀtype
structure) inferred from the XRD results that galliaꢀalumina solid
3
+
53
are assigned to O2s and Ga 3d bands. As for GAꢀHS, a new
and additional lowꢀenergy peak at 19.6 eV is found, which can be
δ+
21, 54
attributed to the presence of Ga species (δ< 2),
suggesting
that the reduced Ga species are probably generated by the 80 solution (cubic spinelꢀtype structure) is generated. In case of GAꢀ
hydrothermal synthesis. This can be further verified by Ga 2p1/2
XPS spectra (Fig. 6B). It can be seen that the GAꢀCP and GAꢀ
GM only show one oxidation state, and Ga 2p1/2 binding energy
HS, besides amorphous spongeꢀlike morphology like GAꢀCP as
shown in Fig. 7d, an additional morphology (Fig. 7e) that some
pieces are covered with amorphous nanoparticles is also seen.
Furthermore, some amorphous nanoparticles can embed into
(BE) value for GAꢀCP (1144.6 eV) is higher than that for GAꢀ
GM (1144.2 eV). While GAꢀHS show two oxidation states 85 these pieces (Fig. 7f). It’s easily understandable that the distinct
(1144.2 eV and 1147.9 eV, respectively). It is deduced that the
other Ga oxidation state (1147.9 eV) ascribed to Ga species
multiple morphologies of these samples are ascribed to their
different structures, which are determined by preparation method
and treatment process.
δ+
probably is formed on GAꢀHS, due to the surface enrichment of
T
Ga species and generation of more lowꢀcoordinated Ga species
3
.5 Acidity of the catalysts
by hydrothermal treatment. The dehydrogenation of propane over
reducible metal oxide catalysts such as chromium and iron oxides
9
0
A comparison of the DRIFT spectra in the OH stretching region
for the asꢀprepared samples is showed in Fig. 8. The stretching
in the presence of CO has been suggested to follow a redox
2
55
band of GaꢀOH groups for Ga O and AlꢀOH for Al O are
mechanism. That is, propane is oxidized to propylene with the
simultaneous reduction of metal oxide (e.g. Fe O , Cr O ), and
2
3
2
3
ꢀ
1
ꢀ1
located at 3650 cm and 3716 cm respectively. As for GAꢀGM,
no any stretching bands of AlꢀOH groups or GaꢀOH groups are
observed but only a broad region, probably because of the
2
3
2
3
subsequently the reduced metal oxide catalyst is reoxidized by
δ+
95
CO . According to the above results, Ga species (δ< 2) exist in
2
formation of disordered mixture between Al O and Ga O . GAꢀ
our prepared GAꢀHS and the mentioned redox mechanism is
probably applicable for the present catalyst system. However,
compared that the fresh catalyst, the spent GAꢀHS with or
2
3
2
3
CP and GAꢀHS both show a new stretching band of special group
ꢀ
1
at about 3683 cm , which is assigned to GaꢀOH or AlꢀOH in the
cubic spinelꢀtype structure. The appearance of the new OH group
without CO (Fig. S7A) shows two binding energy peaks in Ga
2
1
1
1
00 somewhat affects the acidity of catalyst. Based on that, the
distribution of acidity on these different catalysts is measured by
3d XPS spectra at about 24.1 eV and 20.8 eV assigned to O2s and
3
+
Ga 3d bands respectively, and yet the original lowꢀenergy peak
δ+
NH ꢀTPD and the profiles are shown in Fig. 9. It can be seen that
at 19.6 eV attributed to the presence of Ga species (δ< 2)
3
δ+
there is only one NH desorption peak over these catalysts. The
disappears. It is deduced that these Ga species have been
3
3
+
desorption peak on GAꢀHS shows an obvious shift to higher
05 temperature (315 °C) in comparison with that on GAꢀGM
converted into Ga species probably. This can be also further
verified by the disappearance of binding energy peak at 1147.9
eV (ascribed to the Ga oxidation state) in Ga 2p1/2 XPS spectra
(284 °C) and GAꢀCP (303 °C), meaning that the acidity strength
of the GAꢀHS is enhanced facilely via the hydrothermal synthesis
and the mediumꢀstrong Lewis acid sites are generated on the
catalyst accordingly. Moreover, the increased number of Lewis
10 acid sites over GAꢀHS, in comparison of GAꢀCP and GAꢀGM
(
Fig. S7B) after reaction with or without CO . Compared with
2
fresh GAꢀHS, oxidation states in the spent GAꢀHS with CO are
2
similar to that without CO , implying that the redox mechanism in
2
which the CO is taken as an oxidant is not available for the
2
(Table 3), implies that the hydrothermal synthesis promotes the
present catalyst. However, CO still plays a positive role in the
2
formation of mediumꢀstrong Lewis acid sites. It has been proved
that the dehydrogenation of propane proceeded mainly on the
enhancement of dehydrogenation activity through reverse waterꢀ
gas shift reaction and elimination of coke by Boudouard
8
| Catal. Sci. Technol. [2015], [vol], 00–00
This journal is © The Royal Society of Chemistry [2015]

Page 9 of 12
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY02161H
Fig. 8 DRIFT spectra of asꢀprepared samples in the OH stretching region
were recorded after treating the catalyst inꢀsitu at 300 °C for 1 h with a
flow of argon. (a) Ga ; (b) GAꢀGM; (c) GAꢀCP; (d) GAꢀHS; (e) Al
5
2
O
3
2 3
O .
Fig. 11 Transient responses of m/z = 44 (C
catalysts against a pulsed introduction of C
mixture gas (10%CO and 90%Ar). Reaction conditions: catalyst = 100
mg; CO and Ar carrier = 30 mlmin ; C
50°C. (a) GAꢀGM; (b) GAꢀCP; (c) GAꢀHS.
3
H
8
and CO
2
) over asꢀprepared
3
H
8
under steady flow of
15
2
ꢀ1
2
3 8
H = 1 ml; furnace temperature =
5
3+
Lewis acid sites related to lowꢀcoordinated Ga ions on Ga O –
2
3
containing catalysts, which affected intensively their activities for
1
2, 56
2
2
3
3
4
4
0
5
0
5
0
5
dehydrogenation of propane.
It’s apparently plausible that
large amount of Lewis acid sites over GAꢀHS play a key role in
improving the catalytic performance for propane dehydrogenation.
Apart from the acidity number and acidity strength of active
sites over these catalysts, the acidity type also contributes to their
catalytic performance for propane dehydrogenation. Based on
that, FTIR spectra of pyridine adsorption on different catalysts
are recorded to identify the Brønsted acid sites and Lewis acid
sites, and the results are shown in Fig. 10. two peaks, at 1458 cm
and 1620 cm respectively, appear over all these catalysts,
which are assigned to Lewis acid sites.
Lewis acid sites exist on these Ga O –Al O catalysts. In view of
3 2 3 2 3
Fig. 9 (A) NH ꢀTPD profiles of different Ga O –Al O catalysts: (a) GAꢀ
GM; (b) GAꢀCP; (c) GAꢀHS.
ꢀ
1
ꢀ1
44, 57
That’s to say, only
2
3
2
3
NH ꢀTPD profiles, it is deduced that the desorbed NH arise from
3
3
Lewis acid sites. And these Lewis acid sites are related to
3
+
57
coordinately unsaturated Ga ions in the tetrahedral position.
3+
It’s apparent that GAꢀHS possesses more Ga ions in the
tetrahedral sites than GAꢀGM and GAꢀCP because of more
number of Lewis acid sites on GAꢀHS inferred from by NH ꢀTPD,
thereby showing better catalytic performance substantially. This
result is consistent with the NMR results.
reported that surface acidity (specifically
Lewis acid sites density of surface) on the Ga O –Al O catalysts
is crucial to obtain high activity and stability in propane
dehydrogenation. Wherein, the surface acidity is attributed to the
surface tetrahedral Ga ions. Therefore, the investigation to
3
11, 12
Chen et al.
2
3
2
3
3
+
1
0
Fig. 10 PyꢀFTIR spectra of different Ga
2
O
3
–Al
2
O
3
catalysts: (a) GAꢀGM;
3+
surface tetrahedral Ga sites is helpful to understand the propane
(b) GAꢀCP; (c) GAꢀHS.
dehydrogenation inꢀdeep. As proposed, propane activation firstly
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 | 9

Catalysis Science & Technology
Page 10 of 12
View Article Online
DOI: 10.1039/C5CY02161H
3
+
proceeds on these lowꢀcoordinated Ga cations, forming propylꢀ
Ga species, and then the dehydrogenation products are generated
by subsequent decomposition of the resulting propylꢀGa species.
prepared using grindꢀmixture method and GAꢀCP (26.2%)
prepared using coprecipitation method. Furthermore, the
propylene selectivity over GAꢀHS was highest among these
In our work, a positive correlation between the number of Ga 60 catalysts during the whole reaction time. On the other hand,
5
species in tetrahedral sites and the number of surface Lewis acid
sites on GAꢀHS further verifies the previous work.
propane conversion and propylene selectivity as a function of
1
2
time on stream on regenerated GAꢀHS with CO are some low in
2
comparison with that on fresh GAꢀHS, implying that the coke can
be partly eliminated with CO2 through Boudouard reaction.
3
.6 Pulsed reaction
To gain more insight into the mechanism of DHPꢀCO over the 65 Besides, the introduction of CO
to this reaction system improved
durability of the catalysts in spite of the decreased initial
conversions of propane on these Ga ꢀAl catalysts.
2
2
different catalysts, a pulsed reaction technique is used to measure
1
3
10
15
20
25
30
the activities of these catalysts. The transient responses over asꢀ
prepared catalysts against a pulsed introduction of C H at 550°C
2
O
3
O
2 3
Characterization results such as BET, XPS, etc. suggested that
hydrothermal conditions such as generated selfꢀpressure,
3
8
under steady flow of mixture gas (10% CO and 90% Ar) are
2
tested by an onꢀline mass spectrometer as well as the contrast 70 hydrothermal temperature and time (170 °C for 24 h) played
tests, and the results are shown in Fig. 11. When C H is pulsely
significant roles in improving the dispersion and distribution of
3
8
3
+
introduced under steady flow of pure Ar, similar narrow and
positive peaks with the transient response of m/z= 44 (C H ) over
surface gallium sites. This process made more tetrahedral Ga
sites expose on the surface of GAꢀHS, which contributed to
formation of surface Lewis acid sites. Moreover, the
75 hydrothermal treatment is favorable for improving surface area
and increasing pore volume of the catalyst. The texture of GAꢀHS
was largely recovered with air regeneration even after 9 h
reaction time well. Higher surface area and larger amount of
3
8
all these catalysts (Fig. S8) can be seen. Besides, the transient
5
8, 59
response of m/z= 43 determined as C H also can be noticed.
3
8
When steady flow is introduced with 10% CO , the peak
2
responses of m/z = 43 and m/z = 44 over theses catalysts are
obviously different. The transient response of m/z= 43 over GAꢀ
GM, GAꢀCP and GAꢀHS are all similar and present as narrow
3+
δ+
tetrahedral Ga ions (Ga and probably Ga ) related to strong
peaks (Fig. S9), suggesting that C H is detected over these 80 Lewis acid sites were responsible for the superior activity of GAꢀ
3
8
catalysts. In case of the responses of m/z = 44 over GAꢀHS, the
HS catalyst. The correlative work about generation and effect of
Ga ions are still onꢀgoing.
.
δ+
inverted peaks apparently indicate that the introduced C H and
3
8
CO has been consumed greatly (Fig. 11c). Whereas, the narrow
2
and positive peaks of GAꢀCP imply that the remarkable amount
Acknowledgements
of C H feedstock is still reserved (Fig. 11b). The largest broad 85 This research was supported by Key Science and Technology
3
8
and positive peaks of the transient response (m/z = 44) over GAꢀ
GM (Fig. 11a), indicate the weak conversion of C H .
Conclusively, the above results suggest that activity of GAꢀHS
for DHPꢀCO2 is higher than that over other catalysts, which
agrees with the previous test results.
Program of Shanxi Province (No.MD2014ꢀ10), Chinese
Academy of Sciences Knowledge Innovation Project
(No.KGCX2ꢀYWꢀ318ꢀ1) and National Natural Science
Foundation of China (No.21573269)
3
8
60ꢀ63
90
Notes and references
Reported work
has shown that hydrothermal conditions
35
40
45
50
(generated selfꢀpressure, hydrothermal temperature and time) had
great influence on the crystallization and formation process of the
catalysts. In this work, the characterization data suggested that
hydrothermal conditions such as generated selfꢀpressure,
hydrothermal temperature and time (170 °C for 24 h) played
significant roles in improving the dispersion and distribution of
surface gallium sites. Through the hydrothermal process, more
a.
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry,
Chinese Academy of Sciences, Taiyuan 030001, China;
b
University of Chinese Academy of Sciences, Beijing 100049, China;
National Engineering Research Center for CoalꢀBased Synthesis,
c
9
5
Chinese Academy of Sciences, Taiyuan 030001, China.
†
4
Supplementary text; Fig.S1ꢀS9; Table S1ꢀS2. See
DOI: 10.1039/b000000x/
Eꢀmail: tan@sxicc.ac.cn. Tel:+86ꢀ351ꢀ4044287. Fax:+86ꢀ351ꢀ
044287.
3+
tetrahedral Ga sites gallium sites were exposed on the surface of
GAꢀHS, thus contributing to formation of surface Lewis acid sites.
Moreover, this hydrothermal treatment is also favorable for
improving surface area and increasing pore volume of the
catalyst. It’s the superior structure and physical properties
obtained by hydrothermal synthesis that ultimately result in high
activity for dehydrogenation of propane. Therefore, it can be
concluded apparently that hydrothermal synthesis is quite an
effective method to obtain an efficient Ga O ꢀAl O catalyst for
1
00
1
2
.
.
T. Davies and S. H. Taylor, Catal. Lett. , 2004, 93, 151ꢀ154.
M. Chen, J. Xu, Y.ꢀM. Liu, Y. Cao, H.ꢀY. He and J.ꢀH. Zhuang,
Appl. Catal., A, 2010, 377, 35ꢀ41.
3. R. Grabowski, Catal. Rev., 2006, 48, 199ꢀ268.
05 4. D. Vitry, J.ꢀL. Dubois and W. Ueda, J. Mol. Catal. A: Chem. , 2004,
20, 67ꢀ76.
1
2
5
.
K. Takehira, Y. Ohishi, T. Shishido, T. Kawabata, K. Takaki, Q.
Zhang and Y. Wang, J. Catal. , 2004, 224, 404ꢀ416.
2
3
2
3
6. Y. Sakurai, T. Suzaki, N.ꢀo. Ikenaga and T. Suzuki, Appl. Catal., A,
propane dehydrogenation.
110
2000, 192, 281ꢀ288.
7
8
.
.
H. Li, Y. Yue, C. Miao, Z. Xie, W. Hua and Z. Gao, Catal. Commun.
2007, 8, 1317ꢀ1322.
,
4
. Conclusion remarks
Y. Ohishi, T. Kawabata, T. Shishido, K. Takaki, Q. Zhang, Y. Wang
and K. Takehira, J. Mol. Catal. A: Chem. , 2005, 230, 49ꢀ58.
A Ga O –Al O catalyst (GAꢀHS) prepared using hydrothermal
synthesis method showed prominent activity for the DHPꢀCO2.
The highest propane conversion achieved on GAꢀHS catalyst was
2
3
2
3
1
15 9. Y. Wang, Y. Ohishi, T. Shishido, Q. Zhang, W. Yang, Q. Guo, H.
Wan and K. Takehira, J. Catal. , 2003, 220, 347ꢀ357.
10. H. Shimada, T. Akazawa, N.ꢀo. Ikenaga and T. Suzuki, Appl. Catal.,
A, 1998, 168, 243ꢀ250.
55
3
5.2%, which was much higher than that of GAꢀGM (8.7%)
1
0
| Catal. Sci. Technol. [2015], [vol], 00–00
This journal is © The Royal Society of Chemistry [2015]

Page 11 of 12
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY02161H
1
1
1. M. Chen, J. Xu, Y.ꢀM. Liu, Y. Cao, H.ꢀY. He, J.ꢀH. Zhuang and K.ꢀ
N. Fan, Catal. Lett. , 2008, 124, 369ꢀ375.
2. M. Chen, J. Xu, F. Su, Y. Liu, Y. Cao, H. He and K. Fan, J. Catal. ,
43. C. O. Areán, A. L. Bellan, M. P. Mentruit, M. R. Delgado and G. T.
Palomino, Microporous Mesoporous Mater. , 2000, 40, 35ꢀ42.
44. J. Lavalley, M. Daturi, V. Montouillout, G. Clet, C. O. Areán, M. R.
Delgado and A. SahibedꢀDine, PCCP 2003, 5, 1301ꢀ1305.
45. M. Occelli, H. Eckert, A. Wölker and A. Auroux, Microporous
Mesoporous Mater. , 1999, 30, 219ꢀ232.
2
008, 256, 293ꢀ300.
75
5
0
5
0
5
0
5
0
5
0
5
0
5
0
13. K. Nakagawa, C. Kajita, K. Okumura, N.ꢀo. Ikenaga, M. Nishitaniꢀ
Gamo, T. Ando, T. Kobayashi and T. Suzuki, J. Catal. , 2001, 203,
8
7ꢀ93.
46. H. K. C. Timken and E. Oldfield, J. Am. Chem. Soc. , 1987, 109,
7669ꢀ7673.
1
1
4. D. Massiot, T. Vosegaard, N. Magneron, D. Trumeau, V.
Montouillout, P. Berthet, T. Loiseau and B. Bujoli, Solid State Nucl. 80 47. I. Nowak, J. Quartararo, E. G. Derouane and J. C. Védrine, Appl.
Magn. Reson. , 1999, 15, 159ꢀ169.
5. O. Krasnobaeva, I. Belomestnykh, T. Nosova, T. Elizarova, G.
Isagulyants and V. Danilov, Russ. J. Inorg. Chem. , 2013, 58, 644ꢀ
1
1
2
2
3
3
4
4
5
5
6
6
7
Catal., A, 2003, 251, 107ꢀ120.
48. A. Daccà, G. Gemme, L. Mattera and R. Parodi, Appl. Surf. Sci. ,
1998, 126, 219ꢀ230.
49. A. Pijpers and W. Donners, J. Polym. Sci. Polym. Chem., 1985, 23,
453ꢀ462.
50. J. C. Bertolini, J. Massardier, P. Delichere, B. Tardy, B. Imelik, Y.
Jugnet, T. M. Duc, L. De Temmerman, C. Creemers, H. Van Hove
and A. Neyens, Surf. Sci. , 1982, 119, 95ꢀ106.
51. P. Varga and G. Hetzendorf, Surf. Sci. , 1985, 162, 544ꢀ549.
90 52. H. Xiao, J. Zhang, X. Wang, Q. Zhang, H. Xie, Y. Han and Y. Tan,
Catal. Sci. Technol., 2015, 5, 4081ꢀ4090.
53. R. Carli, C. Bianchi, R. Giannantonio and V. Ragaini, J. Mol. Catal. ,
1993, 83, 379ꢀ389.
54. B. M. Reddy, I. Ganesh, E. P. Reddy, A. Fernández and P. G.
6
49.
1
1
1
6. P. Michorczyk and J. Ogonowski, React. Kinet. Catal. Lett. , 2003,
78, 41ꢀ47.
7. P. Michorczyk, K. Zeńczak, R. Niekurzak and J. Ogonowski, Pol. J.
Chem. Technol., 2012, 14, 77ꢀ82.
85
8. T. Davies and S. H. Taylor, J. Mol. Catal. A: Chem. , 2004, 220, 77ꢀ
8
4.
19. K. Nagaoka, K. Seshan, K.ꢀi. Aika and J. A. Lercher, J. Catal. , 2001,
97, 34ꢀ42.
0. P. Meriaudeau, S. A. Hamid and C. Naccache, J. Catal. , 1993, 139,
79ꢀ682.
1
2
2
6
1. S. E. Collins, M. A. Baltanás, J. L. G. Fierro and A. L. Bonivardi, J.
Catal. , 2002, 211, 252ꢀ264.
2. S. Wang and Z. Zhu, Energ. Fuel., 2004, 18, 1126ꢀ1139.
3. Ž. Kotanjac, M. van Sint Annaland and J. Kuipers, Chem. Eng. Sci. ,
95
Smirniotis, J. Phys.Chem.B, 2001, 105, 6227ꢀ6235.
55. P. Michorczyk, P. Kuśtrowski, L. Chmielarz and J. Ogonowski,
React. Kinet. Catal. Lett. , 2004, 82, 121ꢀ130.
56. M. Chen, J. Xu, Y. Cao, H.ꢀY. He, K.ꢀN. Fan and J.ꢀH. Zhuang, J.
Catal. , 2010, 272, 101ꢀ108.
2
2
2
010, 65, 441ꢀ445.
2
2
2
4. S. Tan, S. J. Kim, J. S. Moore, Y. Liu, R. S. Dixit, J. G. Pendergast, 100 57. A. Vimont, J. Lavalley, A. SahibedꢀDine, C. Otero Areán, M.
D. S. Sholl, S. Nair and C. W. Jones, ChemCatChem 2015.
5. B. Xu, T. Li, B. Zheng, W. Hua, Y. Yue and Z. Gao, Catal. Lett. ,
Rodríguez Delgado and M. Daturi, J. Phys.Chem.B, 2005, 109, 9656ꢀ
9664.
2
007, 119, 283ꢀ288.
6. I. Takahara, M. Saito, M. Inaba and K. Murata, Catal. Lett. , 2004,
6, 29ꢀ32.
58. W. A. Chupka and J. Berkowitz, J.Chem.Phys., 1967, 47, 2921ꢀ2933.
59. J. Gilpin and F. McLafferty, Anal. Chem. , 1957, 29, 990ꢀ994.
105 60. O. Yaghi and H. Li, J. Am. Chem. Soc. , 1995, 117, 10401ꢀ10402.
61. X. Wang and Y. Li, J. Am. Chem. Soc. , 2002, 124, 2880ꢀ2881.
62. B. Liu and H. C. Zeng, J. Am. Chem. Soc. , 2003, 125, 4430ꢀ4431.
63. C. S. Cundy and P. A. Cox, Chem. Rev. , 2003, 103, 663ꢀ702.
9
27. M. Saito, S. Watanabe, I. Takahara, M. Inaba and K. Murata, Catal.
Lett. , 2003, 89, 213ꢀ217.
8. K. Nakagawa, C. Kajita, Y. Ide, M. Okamura, S. Kato, H. Kasuya, N.
o. Ikenaga, T. Kobayashi and T. Suzuki, Catal. Lett. , 2000, 64, 215ꢀ
2
29. J.ꢀL. Wu, M. Chen, Y.ꢀM. Liu, Y. Cao, H.ꢀY. He and K.ꢀN. Fan,
Catal. Commun. , 2013, 30, 61ꢀ65.
0. P. Michorczyk, P. Kuśtrowski, A. Kolak and M. Zimowska, Catal.
Commun. , 2013, 35, 95ꢀ100.
1. H. Li, Y. Yue, C. Miao, Z. Xie, W. Hua and Z. Gao, Catal. Commun.
, 2007, 8, 1317ꢀ1322.
2. S. Tan, L. B. Gil, N. Subramanian, D. S. Sholl, S. Nair, C. W. Jones,
J. S. Moore, Y. Liu, R. S. Dixit and J. G. Pendergast, Appl. Catal., A,
2
21.
110
3
3
3
2
015, 498, 167ꢀ175.
3
3
3. P. Michorczyk and J. Ogonowski, Appl. Catal., A, 2003, 251, 425ꢀ
433.
4. J. J. Sattler, I. D. Gonzalez‐Jimenez, L. Luo, B. A. Stears, A.
Malek, D. G. Barton, B. A. Kilos, M. P. Kaminsky, T. W. Verhoeven
and E. J. Koers, Angewandte Chemie, 2014, 126, 9405ꢀ9410.
5. K. Nakagawa, M. Okamura, N. Ikenaga, T. Suzuki and T. Kobayashi,
Chem. Commun. , 1998, 1025ꢀ1026.
3
3
3
6. B. Zheng, W. Hua, Y. Yue and Z. Gao, J. Catal. , 2005, 232, 143ꢀ
1
51.
7. B. Xu, B. Zheng, W. Hua, Y. Yue and Z. Gao, J. Catal. , 2006, 239,
70ꢀ477.
4
38. K.ꢀi. Shimizu, M. Takamatsu, K. Nishi, H. Yoshida, A. Satsuma, T.
Tanaka, S. Yoshida and T. Hattori, J. Phys. Chem. B, 1999, 103,
1
9. V. B. Kazansky, I. R. Subbotina, A. A. Pronin, R. Schlögl and F. C.
Jentoft, J. Phy. Chem. B, 2006, 110, 7975ꢀ7978.
40. C. O. Areán, M. R. Delgado, V. Montouillout and D. Massiot, Z.
Anorg. Allg. Chem. , 2005, 631, 2121ꢀ2126.
1. J. Yu, C. Y. Jimmy, M. K.ꢀP. Leung, W. Ho, B. Cheng, X. Zhao and
J. Zhao, J. Catal. , 2003, 217, 69ꢀ78.
2. F. Gennari, A. C. Ramos, A. Condó, T. Montini, S. Bengió, A.
Cortesi, J. A. Gamboa and P. Fornasiero, Appl. Catal., A, 2011, 398,
542ꢀ1549.
3
4
4
1
23ꢀ133.
This journal is © The Royal Society of Chemistry [year]
Catal. Sci. Technol. [2015], [vol], 00–00 | 11

Catalysis Science & Technology
Page 12 of 12
View Article Online
DOI: 10.1039/C5CY02161H
Hydrothermal synthesized Ga O –Al O catalysts showed superior
2
3
2
3
activity for dehydrogenation of propane in the presence of carbon
dioxide